本文为《Linear algebra and its applications》的读书笔记
目录
- Difference equations are often the appropriate tool to analyze discrete, or digital data. Even when a differential equation is used to model a continuous process, a numerical solution is often produced from a related difference equation.
- This section highlights some fundamental properties of linear difference equations.
Discrete-Time Signals
离散时间信号
- The vector space
S
\mathbb S
S of discrete-time signals was introduced in Section 4.1. A signal in
S
\mathbb S
S is a function defined only on the integers and is visualized as a sequence of numbers, say,
{
y
k
}
\{y_k\}
{yk}.
Linear Independence in the Space S \mathbb S S of Signals
- To simplify notation, we consider a set of only three signals in
S
\mathbb S
S, say,
{
u
k
}
\{u_k\}
{uk},
{
v
k
}
\{v_k\}
{vk}, and
{
w
k
}
\{w_k\}
{wk}. They are linearly independent precisely when the equation
c 1 u k + c 2 v k + c 3 w k = 0 f o r a l l k ( 1 ) c_1u_k + c_2v_k +c_3w_k = 0\ \ \ \ \ \ \ for\ all\ k\ \ \ \ \ \ \ \ \ \ \ \ (1) c1uk+c2vk+c3wk=0 for all k (1)implies that c 1 = c 2 = c 3 = 0 c_1 = c_2 = c_3 = 0 c1=c2=c3=0. - Suppose
c
1
,
c
2
,
c
3
c_1, c_2, c_3
c1,c2,c3 satisfy (1). Then equation (1) holds for any three consecutive values of
k
k
k, say,
k
k
k,
k
+
1
k + 1
k+1, and
k
+
2
k +2
k+2. Thus (1) implies that
The coefficient matrix in this system is called the Casorati matrix of the signals, and the determinant of the matrix is called the Casoratian of { u k } \{u_k\} {uk}, { v k } \{v_k\} {vk}, and { w k } \{w_k\} {wk}.
- If the Casorati matrix is invertible for at least one value of k k k, then (2) will imply that c 1 = c 2 = c 3 = 0 c_1 = c_2 = c_3 = 0 c1=c2=c3=0, which will prove that the three signals are linearly independent.
- If a Casorati matrix is not invertible, the associated signals being tested may or may not be linearly dependent.
- However, it can be shown that if the signals are all solutions of the same homogeneous difference equation, then either the Casorati matrix is invertible for all k k k and the signals are linearly independent, or else the Casorati matrix is not invertible for all k k k and the signals are linearly dependent. (Check the appendix)
EXAMPLE 2
Verify that 1 k , ( − 2 ) k 1^k, (-2)^k 1k,(−2)k, and 3 k 3^k 3k are linearly independent signals.
SOLUTION
- The Casorati matrix is
Row operations can show fairly easily that this matrix is always invertible. However, it is faster to substitute a value for k k k—say, k = 0 k = 0 k=0—and row reduce the numerical matrix:
So 1 k , ( − 2 ) k 1^k, (-2)^k 1k,(−2)k, and 3 k 3^k 3k are linearly independent signals.
Linear Difference Equations
- Given scalars
a
0
,
.
.
.
,
a
n
a_0,..., a_n
a0,...,an, with
a
0
a_0
a0 and
a
n
a_n
an nonzero, and given a signal
{
z
k
}
\{z_k\}
{zk}, the equation
is called a linear difference equation (or linear recurrence relation) of order n \boldsymbol n n ( n \boldsymbol n n 阶线性差分方程 ). For simplicity, a 0 a_0 a0 is often taken equal to 1 1 1. If { z k } \{z_k\} {zk} is the zero sequence, the equation is homogeneous; otherwise, the equation is nonhomogeneous.
y k + 3 + 5 y k + 2 + 6 y k + 1 = 0 y_{k+3}+5y_{k+2}+6y_{k+1}=0 yk+3+5yk+2+6yk+1=0 是 2 2 2 阶差分方程 (将 k k k 用 k − 1 k-1 k−1 替换)
EXAMPLE 3
- In digital signal processing, a difference equation such as (3) describes a linear filter (线性滤波器), and a 0 , . . . , a n a_0,..., a_n a0,...,an are called the filter coefficients. If { y k } \{y_k\} {yk} is treated as the input and { z k } \{z_k\} {zk} as the output, then the solutions of the associated homogeneous equation are the signals that are filtered out and transformed into the zero signal.
- Let us feed two different signals into the filter
Here . 35 .35 .35 is an abbreviation for 2 / 4 \sqrt 2/4 2/4.
- The first signal is created by sampling the continuous signal
y
=
cos
(
π
t
/
4
)
y = \cos(\pi t/4)
y=cos(πt/4) at integer values of
t
t
t, as in Figure 3(a). The discrete signal is
For simplicity, write ± . 7 \pm .7 ±.7 in place of ± 2 / 2 \pm \sqrt 2/2 ±2/2, so that
Table 1 shows a calculation of the output sequence { z k } \{z_k\} {zk}. The output is { y k } \{y_k\} {yk} shifted by one term.
- A different input signal is produced from the higher frequency signal
y
=
cos
(
3
π
t
/
4
)
y = \cos(3\pi t/4)
y=cos(3πt/4), shown in Figure 3(b). Sampling at the same rate as before produces a new input sequence:
When { w k } \{w_k\} {wk} is fed into the filter, the output is the zero sequence. The filter, called a l o w − p a s s low-pass low−pass f i l t e r filter filter, lets { y k } \{y_k\} {yk} pass through, but stops the higher frequency { w k } \{w_k\} {wk}.
- The first signal is created by sampling the continuous signal
y
=
cos
(
π
t
/
4
)
y = \cos(\pi t/4)
y=cos(πt/4) at integer values of
t
t
t, as in Figure 3(a). The discrete signal is
- In many applications, a sequence { z k } \{z_k\} {zk} is specified for the right side of a difference equation (3), and a { y k } \{y_k\} {yk} that satisfies (3) is called a solution of the equation. The next example shows how to find solutions for a homogeneous equation.
EXAMPLE 4
Solutions of a homogeneous difference equation often have the form y k = r k y_k = r^k yk=rk for some r r r. Find some solutions of the equation
SOLUTION
- Substitute
r
k
r^k
rk for
y
k
y_k
yk in the equation and factor the left side:
Thus 1 k 1^k 1k, ( − 2 ) k (-2)^k (−2)k, and 3 k 3^k 3k are all solutions of (4).
- In general, a nonzero signal
r
k
r^k
rk satisfies the homogeneous difference equation
if and only if r r r is a root of the auxiliary equation
When the auxiliary equation has a c o m p l e x r o o t complex\ root complex root(复数根), the difference equation has solutions of the form s k cos k w s^k \cos kw skcoskw and s k sin k w s^k \sin kw sksinkw, for constants s s s and w w w.
Solution Sets of Linear Difference Equations
- Given
a
1
,
.
.
.
,
a
n
a_1,..., a_n
a1,...,an, consider the mapping
T
:
S
→
S
T: \mathbb S \rightarrow \mathbb S
T:S→S that transforms a signal
{
y
k
}
\{y_k\}
{yk} into a signal
{
w
k
}
\{w_k\}
{wk} given by
It is readily checked that T T T is a l i n e a r linear linear transformation. This implies that the solution set of the homogeneous equation
is the kernel of T T T, and hence the solution set is a s u b s p a c e subspace subspace of S \mathbb S S. Any linear combination of solutions is again a solution.
- The next theorem, a simple but basic result, will lead to more information about the solution sets of difference equations.
PROOF
- If
y
0
,
.
.
.
,
y
n
−
1
y_0,..., y_{n-1}
y0,...,yn−1 are specified, use (7) to define
And now that y 1 , . . . , y n y_1,..., y_{n} y1,...,yn are specified, use (7) to define y n + 1 y_{n+1} yn+1. In general, use the recurrence relation
to define y n + k y_{n+k} yn+k for k ≥ 0 k \geq 0 k≥0. To define y k y_k yk for k < 0 k < 0 k<0, use the recurrence relation
So the solution of (7) is unique.
PROOF
- As was pointed out earlier, H H H is a subspace of S \mathbb S S because H H H is the kernel of a linear transformation.
- For { y k } \{y_k\} {yk} in H H H, let F { y k } F\{y_k\} F{yk} be the vector in R n \mathbb R^n Rn given by ( y 0 , y 1 , . . . , y n − 1 ) (y_0, y_1,..., y_{n-1}) (y0,y1,...,yn−1). It is readily verified that F : H → R n F: H \rightarrow \mathbb R^n F:H→Rn is a linear transformation. Given any vector ( y 0 , y 1 , . . . , y n − 1 ) (y_0, y_1,..., y_{n-1}) (y0,y1,...,yn−1) in R n \mathbb R^n Rn, Theorem 16 says that there is a unique signal { y k } \{y_k\} {yk} in H H H such that F { y k } = ( y 0 , y 1 , . . . , y n − 1 ) F \{y_k\}= (y_0, y_1,..., y_{n-1}) F{yk}=(y0,y1,...,yn−1). This means that F F F is a one-to-one linear transformation of H H H onto R n \mathbb R^n Rn; that is, F F F is an isomorphism. Thus d i m H = d i m R n = n dim\ H = dim\ \mathbb R^n = n dim H=dim Rn=n.
EXAMPLE 5
Find a basis for the set of all solutions to the difference equation
SOLUTION
- We know from Examples 2 and 4 that 1 k 1^k 1k, ( − 2 ) k (-2)^k (−2)k, and 3 k 3^k 3k are linearly independent solutions. Theorem 17 shows that the solution space is exactly three-dimensional. So 1 k 1^k 1k, ( − 2 ) k (-2)^k (−2)k, and 3 k 3^k 3k form a basis for the solution space.
- The standard way to describe the “general solution” of the difference equation (10) is to exhibit a basis for the subspace of all solutions. Such a basis is usually called a fundamental set of solutions (基础解系) of (10).
Nonhomogeneous Equations
- The general solution of the nonhomogeneous difference equation
can be written as one particular solution of (11) plus an arbitrary linear combination of a fundamental set of solutions of the corresponding homogeneous equation (10).
- This fact is analogous to the result in Section 1.5 showing that the solution sets of A x = b A\boldsymbol x =\boldsymbol b Ax=b and A x = 0 A\boldsymbol x =\boldsymbol 0 Ax=0 are parallel.
Reduction to Systems of First-Order Equations
化简为一阶方程组
- A modern way to study a homogeneous
n
n
nth-order linear difference equation is to replace it by an equivalent system of first-order difference equations, written in the form
where the vectors x k \boldsymbol x_k xk are in R n \mathbb R^n Rn and A A A is an n × n n \times n n×n matrix.
EXAMPLE 7
Write the following difference equation as a first-order system:
SOLUTION
- For each
k
k
k, set
So
That is,
- In general, the equation
can be rewritten as x k + 1 = A x k \boldsymbol x_{k+1} = A\boldsymbol x_k xk+1=Axk for all k k k, where
Appendix: The Casorati Test
Let { y 1 , … , y n } \{\boldsymbol y_1, …, \boldsymbol y_n\} {y1,…,yn} be a set of signals in S \mathbb S S. For j = 1 , … , n j = 1, …, n j=1,…,n and for any k k k, let y j ( k ) \boldsymbol y_j(k) yj(k) denote the k k kth entry in the signal y j \boldsymbol y_j yj and let
- a. If C ( k ) C(k) C(k) is invertible for some k k k, { y l , … , y n } \{\boldsymbol y_l, …, \boldsymbol y_n\} {yl,…,yn} is linearly independent.
- b. If
y
l
,
…
,
y
n
\boldsymbol y_l, …, \boldsymbol y_n
yl,…,yn all satisfy a homogeneous difference equation of order
n
n
n,
(with a n ≠ 0 a_n \neq 0 an=0), and if the Casorati matrix C ( k ) C(k) C(k) is not invertible for some k k k, then { y 1 , … , y n } \{\boldsymbol y_1, …, \boldsymbol y_n\} {y1,…,yn} is linearly dependent in S \mathbb S S, and for all k k k, C ( k ) C(k) C(k) is not invertible.
Proof
- (b) Suppose that
y
1
,
…
,
y
n
\boldsymbol y_1, …, \boldsymbol y_n
y1,…,yn are in the set
H
H
H of solutions of
(
∗
)
(*)
(∗) and
C
(
k
0
)
C(k_0)
C(k0) is not invertible for some
k
0
k_0
k0. It is readily verified that if
T
:
H
→
R
n
T: H \rightarrow R^n
T:H→Rn is defined by
then T T T is a linear transformation. The proof of Theorem 16 is easily modified to show that ( ∗ ) (*) (∗) has a unique solution y \boldsymbol y y whenever y ( k 0 ) , … , y ( k 0 + n – 1 ) \boldsymbol y(k_0), …, \boldsymbol y(k_0 + n – 1) y(k0),…,y(k0+n–1) are specified. This means that T T T is a one-to-one mapping of H H H onto R n \mathbb R^n Rn. Furthermore, the images T ( y 1 ) , … , T ( y n ) T(\boldsymbol y_1), …, T(\boldsymbol y_n) T(y1),…,T(yn) form the columns of the Casorati matrix C ( k 0 ) C(k_0) C(k0) and hence are linearly dependent, because C ( k 0 ) C(k_0) C(k0) is not invertible. Since T T T is one-to-one, { y 1 , … , y n } \{\boldsymbol y_1, …, \boldsymbol y_n\} {y1,…,yn} is linearly dependent. This proves the first statement in (b). The second statement follows immediately from part (a), because if C ( k ) C(k) C(k) were invertible for some k k k, then { y 1 , … , y n } \{\boldsymbol y_1, …, \boldsymbol y_n\} {y1,…,yn} would be linearly independent, which is not true. So C ( k ) C(k) C(k) is not invertible for each k k k.